Employment, computer systems design and related servicesbetween
2,355 thousands
on track <= 2,340 · off >= 2,380bls ↗
series as of 2026-09-30
Radar · Enterprise absorption · T2 · 2029 · WARNING
IT services payrolls fall below 2.3 million by 2029
BLS payroll employment in computer systems design and related services (NAICS 5415, series CES6054150001, seasonally adjusted) prints at or below 2,300 thousand in at least one month with a reference period on or before 2029-11-30, about 6% below its 2024 peak, while BLS projects the industry to grow 15.8% over the decade.
ClaimBLS payroll employment in computer systems design and related services (NAICS 5415, series CES6054150001, seasonally adjusted) prints at or below 2,300 thousand in at least one month with a reference period on or before 2029-11-30, about 6% below its 2024 peak, while BLS projects the industry to grow 15.8% over the decade.
Consensus (implied)25%implied from BLS Employment Projections, industries with the largest wage and salary employment growth, 2024-34 · 2025-08-28
Distance+1.95log-odds · far above consensus
My confidence70%80% CI 55–82%
Engine92%+22 pts vs me · stacked-fixed-weights
Falsifies ifCES6054150001 stays above 2,300 thousand in every month through November 2029, or turns up to regain its 2024 peak of about 2,443 thousand.
HorizonDecember 31, 20291169 days · by end-2029 · milestone ladder
Why it matters
IT services is the first large white-collar industry where AI does the work the workers were hired to do: writing, testing, and maintaining code and systems. BLS still projects it as the second-fastest source of new jobs in the country. If payrolls keep falling instead, the labor-market effect of coding agents shows up in a monthly federal series two years before it shows up in any company's 10-K.
Probability over time
Registered at 70% on September 8, 2026. Engine repriced 14 times; now 92%.
Milestone ladder
Dated rungs. Each is scored on its own; the thesis does not get credit for the ladder until the rungs land.
filled bar · my probabilityhollow dot · engineamber date · due, awaiting adjudication
m1 · 2028-01-31 · 70% · CES6054150001 prints at or below 2,340 thousand for any month through December 2027.
m2 · 2029-01-31 · 65% · CES6054150001 prints at or below 2,320 thousand for any month through December 2028.
Leading indicators
Registered thresholds. Status is computed from the latest public reading.
What would move me
Would raise my number
- Three consecutive monthly declines of 5 thousand or more in CES6054150001.
- Large IT outsourcers reporting revenue growth with flat or falling headcount for two fiscal years.
- Indeed software development postings falling below 60 on the February 2020 index.
Would cut it
- Two consecutive months of net gains above 5 thousand in CES6054150001.
- An annual BLS benchmark revision that lifts the 2025-2026 level by more than 30 thousand.
- Enterprise AI deployment creating net new integration and systems work that outsourcers hire for.
Council · 72%
6 personas on one model (openai/gpt-5.6-sol); other vendors were unavailable this run, so read this as one model's persona-averaged view.
67%
grid-and-power-planner · openai/gpt-5.6-sol
September 2026 is only 55.4 thousand above the trigger. The series has fallen roughly 87 thousand since its 2024 peak and 14.1 thousand since April; even half that recent pace would cross before the horizon. Software-development postings remain weak, although productivity data do not establish rapid labor substitution. My reference class is three-year windows for cyclical professional-services payrolls already in a multiyear decline: an annual crossing hazard near 0.28 compounds to about 0.65, modestly raised for proximity and momentum. I discount the 0.98 drift prior because it ignores trend reversals. BLS’s decade projection is an endpoint trend, not protection against an interim trough. First-print and discontinuation provisions trim the estimate.
72%
semiconductor-supply-chain-analyst · openai/gpt-5.6-sol
At 2,355.4 in September 2026, the series needs only a further 55.4-thousand, or 2.4%, drawdown over 38 reference months. It fell 14.1 thousand from April through September, while software-development postings near 78 indicate weak hiring. Using cyclical U.S. professional/business-services payroll contractions as the reference class, I assign conditional hit hazards around 0.42 through 2027, 0.32 in 2028, and 0.25 through November 2029. Compounding and first-print volatility support 0.72. The any-month rule helps, but normal hiring recoveries and BLS’s positive long-run projection make the engine’s straight-drift 0.98 prior too high.
68%
enterprise-cio · openai/gpt-5.6-sol
The series is only 55.4 thousand above the trigger after a 25-month, roughly 87-thousand decline; recent first prints show renewed deterioration, and weak software-development postings offer little evidence of a near-term hiring rebound. Using post-peak contractions in cyclical professional-services payrolls as the reference class, I estimate conditional hit hazards near 0.08 for late 2026, 0.42 in 2027, 0.25 in 2028, and 0.20 through November 2029. Compounding supports about 0.68. BLS’s decade projection is a weak short-horizon timing guide, but modest productivity growth and eventual AI implementation demand argue against the engine’s near-certain extrapolation. First-print-only scoring adds some miss risk.
82%
credit-analyst · openai/gpt-5.6-sol
Starting at 2,355.4, the barrier is only 55.4 thousand (2.35%) lower. Payrolls are down about 87 thousand from the 2024 peak, including 14.1 thousand since April and four consecutive monthly declines. At the recent pace, crossing occurs around early 2028, leaving substantial cushion. Weak software-development postings corroborate limited hiring demand; 1.4% productivity growth is not a strong displacement signal. My reference class is cyclical three-year drawdowns in detailed service-industry payrolls: an approximately 0.42 annual crossing hazard over 38 months compounds near 0.82. I discount the 0.98 drift prior for mean reversion and first-print noise; BLS’s decade projection does not preclude a temporary breach.
76%
geopolitics-policy-analyst · openai/gpt-5.6-sol
The series is only 55.4 thousand above the barrier after a roughly 87 thousand decline from its 2024 peak. The latest six-month path is down 14.1 thousand, while software-development postings remain weak; continuation at even a fraction of the recent pace would cross before 2029. I use the reference class of post-peak CES industry contractions plus a 0.15–0.20 annual recession/downturn hazard, compounded over just over three years. Against this, long employment declines often reverse, productivity growth is modest, and BLS projects strong secular industry growth. Those considerations make the engine’s 0.98 drift extrapolation substantially too high.
66%
superforecaster-statistician · openai/gpt-5.6-sol
Base rate: I use a 0.15 annual U.S. recession-start hazard, implying 0.39 over the roughly 3.2-year window; such downturns commonly produce a decline exceeding the required 2.35%. The series is only 55.4 thousand above the trigger and has fallen 87.2 thousand since its 2024 peak, with declines resuming after a spring plateau. Weak software-development postings support continued softness. Against this, BLS projects strong decade growth, productivity gains are modest, and prolonged declines can reverse. I heavily discount the 0.98 drift prior because it extrapolates one regime. Multiple monthly opportunities lift the odds; first-print and continuity clauses trim them.
Engine prior
98% from drift-gbm-terminal:li-1. P(indicator li-1 satisfies onTrack <= 2340.0 at 2029-12-31); drift -0.0143/yr, vol 0.00584/sqrt(yr) from 33 points; proxy for the thesis, not its rule